Octopus Saving Session

topdave

I've spent several hours working out why my saving was only around 50p, despite being a heavy electricity user, and switching off all my appliances and leaving the house. It comes down to the "in-day adjustment". This is used to modify your "average usage", and it's the value they compare to what you actually used.

I can't post document link (I'm a noob), but it's called "Demand Flexibility Service, Participation Guidance Document V.3, November 2022". Accessible from national grid website. See appendix 4.
It's not super clear though. I would be interested in other people's analysis / interpretation.

On the 15th Nov, between 5pm and 6pm, they measured usage. This is compared to your hourly average over the previous 10 week days (unadjusted baseline value), modified by the "in-day adjustment", which is based on your consumption between 1pm and 4pm on the 15th Nov. 
They compared your usage between 5pm and 6pm to (average over 10 days) + (average over 3 hour window - average over 10 days) / 2
ie usage was compared to (average over 10 days + average of 3 hour window) / 2 

My average usage is 1kw, across the 10 days.
Between 1pm and 4pm I used average of 0.5kw
Between 5pm and 6pm I also used 0.5kw
(1kw + 0.5kw) / 2 = 0.75kw "baseline value" (why not call this the adjusted baseline value!)

Implied saving (rate reduction over the hour) = 0.75kw - 0.5kw = 0.25kw
Payment = £2.25/kwh * 0.25kw * 1 hour = 56p
[1 kwh = 1 kilowatt hour = 1 unit = a rate of 1kw over an entire hour]

My average usage in the evening 5pm to 6pm is around 1.5kwh for reference, so I was expecting to be paid closer to £2.25 for my reduction. No, that's not how the calculation works apparently.

Anyway, next time, I'm putting my electric car on charge between 1 and 4 hours before the time period. That will add 3kw to my rate. Then my payment will be
1kw (unadjusted baseline value) + (1kw usual + 3kw from car charging) (ie 4kw in-day adjustment) / 2 = 2.5kw ("baseline value")
Implied saving (over hour) = 2.5kw - 0.5kw = 2kw "saving"
Payment = £2.25/kwh * 2kw * 1 hour = £4.50

Or charge car during in-day adjustment window (1 to 4 hours prior), and make no effort to save electricity during window
2.5kw - 1.5kw = 1kw "saving"
Payment = £2.25/kwh * 1kw * 1 hour = £2.25

Maybe I'm interpreting it all wrong, but it kind of makes sense from a National Grid perspective, they want to pay people to reduce usage relative to what they were earlier in previous hours, as grid want to reduce the current consumption. But that does ignore fact that usage spikes at particular times, and you want to reduce that spike.

Also given price cap, you're getting paid to use electricity during the 3 hour "in-day adjustment" window. 
Say an extra from charging car for 3 hours on normal plug socket
3kw x 3 hours = 9kwh = 9 units
34p/unit * 9 units = £3.06 of electricity
This would increase your "baseline value" by 1.5kw
If during window you just use your average energy, you'd get a payment of
Payment = £2.25/kwh * 1.5kw * 1 hour = £3.38 session payment
This might scheme might need a small rethink, but it's the right thing to be doing to smooth out electricity consumption.
If the schema (payment) "event window" is greater than 1 hour, it would magnify this effect.
I'm also not sure what sort of "capping" is in place to prevent this unintended consequence.

[Deleted User]

If you look back at pages 3 through 8 of this thread, we spent tens of posts discussing the mathematics presented in the documentation.

You're almost right with how you have interpreted it, but not quite (your in-day adjustment calculation is likely off unless you have very strange usage).

The inferences you draw, however, are pretty much correct.

topdave

Does anyone have spreadsheet with exact workings?

[Deleted User]

I've read the DFS Procurement Rules v1 - 27 Oct 2022 over and over.

It just doesn't make sense, if the "inday" adjustment is using data up to 1 hour before the event YET all periods are treated as individual half hour windows then if an event is longer than 1 hour the maths just falls apart.
I'm thinking the supposedly well though out formula might have lost some meaning in translation into words..

I was paid £4.75 for a 95% reduction and awarded and extra £1.25 for been in the top 10%

I have a solar inverter that logs all my energy usage whether from the panels or not and I have no idea how they arrived at that figure, even after reading and digesting the above mentioned document.

Oh well, see what the next session brings I suppose.

[Deleted User]

I suppose its this sentence that I got wrong.  "The In Day Adjustment will consider actual Metered data over the three hour period up until one hour ahead of the relevant Settlement Period when the service starts to deliver."  DOH!

[Deleted User]

Not quite. Hard to explain in words so I’ll try to make a spreadsheet.

Irratus_Rusticus

Haven't a clue how it works, but happy to take part. I saved 37p ! Can't save what I don't use and that was a 69% reduction for the hour. Just router and freezer left on - and the IHD.

If 20,000 Octopus users can save equivalent of turning off an "entire gas power station" in one hour, presumably big guns like BG will save a shed-load more and avoid us facing those blackout levels?  Stocked up on tea candles for next sessions, so no net savings, self.

[Deleted User]

[Deleted_User] and @topdave

I'll try and explain here how the equation can be read - see if you can follow it with the numbers that you recorded.  It's all about getting the right period for the source of each bit of maths.

Each day is broken down into 48 half-hour periods (settlement periods).

Your baseline for each of these 48 periods is calculated using the average of that period in the last ten days.

On the day of a saving session, an adjustment is calculated.  This calculation is done by taking 6 consecutive periods, with the sixth period ending one hour (two periods) before the start of the saving session.

The actual use in each of these 6 periods on the day is compared to the baseline for each of those 6 periods.  Six subtractions, giving you six differences.  These differences are averaged to give you the in-day adjustment (now a fixed number for the day).

For each period in the saving session, however many there are, the fixed in-day adjustment is added to your baseline for that period to give you the adjusted baseline against which your saving will be measured.

tlcgrantham

Next session tomorrow at 5.30 pm for 1 hour

Alnat1

Not sure I'll be able to persuade hubby to turn off the TV if the footy is on.......